Molecular characterization of resistance mechanisms in Pseudomonas aeruginosa isolates resistant to carbapenems

Introduction: Emergence of carbapenem resistance in Pseudomonas aeruginosa increases the therapeutic dilemma. In this study, we investigated various mechanisms involved in the resistance of P. aeruginosa clinical isolates to carbapenems. Methodology: P. aeruginosa isolates were isolated from different clinical samples. The antimicrobial susceptibility was evaluated by disc diffusion method. Carbapenemases were detected among carbapenem resistant isolates. Expression level of mexB and oprD was determined by real-time PCR. Molecular relatedness among isolates was detected based on pulsed-field gel electrophoresis (PFGE). Results: Ninety P. aeruginosa isolates were purified from clinical specimens. High levels of resistance to imipenem and meropenem were detected in 16 isolates. PCR analysis of carbapenemases indicated the prevalence of Verona integron-encoded metallo-beta-lactamase (VIM); four isolates produced only VIM enzymes (VIM-1 or VIM-2), while the remaining twelve co-produced both VIM-1 or VIM-2 and NDM enzymes. Additionally, real-time PCR analysis elucidated high expression levels of mexB in seven of the carbapenem resistant isolates and low expression of oprD in seven isolates. The identified carbapenem-resistant isolates were clustered into eleven PFGE profiles where clusters E1 and E2 involved isolates exhibiting multiple carbapenemase genes (blaNDM-1, blaVIM-1 and blaVIM-2). Conclusion: Various mechanisms underlying carbapenem resistance have been detected in our P. aeruginosa cohort of isolates. Emergence of P. aeruginosa as a reservoir of multiple carbapenemases is increasing over time limiting the treatment options to this serious infection. This increases the urgency for infection control practices to reduce the incidence of this infection.


Introduction
Pseudomonas aeruginosa is one of the most important opportunistic pathogens that has been associated with community and hospital-acquired infections such as respiratory tract infections, burns, wounds, otitis media and nosocomial infections [1]. The frequency of infections caused by P. aeruginosa is increasing and multidrug-resistant (MDR) isolates are emerging in hospitalized patients [2].
Carbapenems have been kept as a last resort therapy for the control of MDR P. aeruginosa infection. However, carbapenem-resistant P. aeruginosa could be resistant to other classes of antimicrobial agents and such infections are associated with limited therapeutic options and high rates of mortality and morbidity especially in hospitalized and immunocompromised patients [3][4].
Various mechanisms contribute to the development of carbapenem-resistant P. aeruginosa particularly through the acquisition of resistant genes encoding carbapenem-hydrolyzing enzymes [5][6]. A wide range of metallo-β-lactamases (MBLs) (class B β-lactamases) such as IMP, VIM, NDM-1 and GIM-1 have been reported in P. aeruginosa. These enzymes play a major role in the resistance of P. aeruginosa to carbapenems [7]. In addition, class A β-lactamase Klebsiella pneumoniae carbapenemases (KPC) carries the extended spectrum KPC-1 enzyme, which was first detected in an outbreak in North Carolina [8], has subsequently been identified in P. aeruginosa isolates [9]. Furthermore, P. aeruginosa isolates may exhibit multidrug efflux machinery, mainly MexAB-OprM and MexXY-OprM. Efflux of carbapenems represents a principal resistance mechanism in P. aeruginosa [10].
The transport of carbapenems across the lipid bilayer membranes occurs through protein channels, porins, especially OprD [11]. Lack of OprD participates in the resistance of P. aeruginosa isolates to carbapenems [12][13]. The aim of this study is to examine the coexistence of different genes encoding carbapenemases in P. aeruginosa clinical isolates. Also, the expression level of porins (OprD) and efflux machinery were evaluated as contributing mechanisms of carbapenem resistance in these isolates. Moreover, the genotypic profile of P. aeruginosa clinical isolates was determined using pulsed-field gel electrophoresis (PFGE). This can shed more light on promising targets for developing strategies to defend the dissemination of carbapenem resistance, propose infection control regimens and assign suitable treatment for such infection.

P. aeruginosa clinical isolates
Gram-negative bacilli were recovered from various sources, including sputum, urine and pus exudates between May 2015 and September 2015. Ninety P. aeruginosa isolates were identified according to standard laboratory methods [14]. P. aeruginosa isolates were collected during outpatient treatment at University Hospital (255 bed), Chest Hospital (214 bed), Urology and Nephrology Center (270 bed), International Hospital (400 bed) and from inpatient admitted Burns and Cosmetics Center (48 bed) at this period. Luria-Bertani medium was used for propagation of bacterial growth. Glycerol stocks (30%) were prepared and stored at -80 ºC. The methodology applied in this research follow the ethical guidelines of "The Research Ethics Committee, Faculty of Pharmacy, Mansoura University".

Assay of antimicrobial susceptibility
The susceptibility of the purified isolates to different types of antimicrobial agents was evaluated using disk diffusion method according to the guidelines of the Clinical and Laboratory Standards Institute (CLSI, 2015) [15]. The tested antimicrobial agents include ceftazidime (CAZ), imipenem (IPM), meropenem (MEM), amikacin (AK), gentamicin (CN), levofloxacin (LEV) (Oxoid, Basingstoke, UK). The susceptibility profiles were interpreted using the Clinical and Laboratory Standards outlines [15].

Determination of the MICs
Minimum inhibitory concentrations (MICs) of imipenem and meropenem against our isolates were determined by microtitre plate broth dilution method. Two-fold serial dilutions of imipenem or meropenem were prepared in Mueller-Hinton broth as follows 2048, 1024, 512, 256, 128, 64, 32, 16, 8, 4, 2, 1 µg per mL. The dilutions were inoculated with 0.5 × 10 5 colonyforming units per mL of the tested isolates and the plates were incubated for 18 hours at 37 °C [15]. Bacterial growth was assessed and the MIC for each isolate was quantified as the lowest concentration of the antibiotic that prevents visible bacterial growth.

Modified Hodge Test (MHT)
The presence of carbapenemases in P. aeruginosa isolates was primary detected using Modified Hodge test. The diluted culture of Escherichia coli ATCC 25922 (0.5 McFarland standard) was swabbed on the surface of Mueller-Hinton agar plates in three different directions. Meropenem disk (10 µg) (Oxoid, Basingstoke, UK) was placed at the center of each plate. The tested isolates were streaked as a thin line from the edge of the meropenem disk to the edge of the plate [16]. Bacterial growth was allowed for 18 hours at 37 °C. Indentation in the inhibition zone of E. coli or clove growth of E. coli around the meropenem disk revealed a positive MHT [15].

PCR of carbapenemases encoding genes
The carbapenemases encoding genes; blaNDM-1, blaVIM-1, blaVIM-2, blaOXA, blaIMP, blaKPC were screened by PCR using 0.5 U FlexiTaq DNA polymerase (Promega, Madison, USA), 5 × GoTaq Flexi buffer, 0.5 mM of each primer (Table 1), 0.1 mM of each dNTP, 0.5 mM MgCl2 solution and H2O up to 20 µL. The thermocycling conditions started with initial denaturation at 95 °C for 5 minutes, followed by 35 cycles of denaturation at 95 °C for 30 seconds, annealing for 30 seconds at temperatures specific to each primer pair as mentioned in Table 1 and extension at 72°C for 30 seconds and the reaction ended with a final extension at 72 °C for 5 minutes.

The role of efflux machinery in carbapenem resistance
The sensitivity of the tested isolates to imipenem was established via microtitre plate broth dilution method with and without the addition of efflux pump inhibitors; carbonyl cyanide m-chlorophenyl hydrazone (CCCP; Sigma-aldrich, St. Louis, USA) at a concentration of 40 µg/mL and Phe-Arg-bnaphthylamide (PAbN; PAbN; Sigma-aldrich, St. Louis, USA) at a concentration of 50 µg/mL [17][18]. The MIC of the imipenem/inhibitor was compared to the MIC of imipenem alone using paired T test with p < 0.01 was considered significant.

Real-time PCR of porin (OprD) and efflux pump
P. aeruginosa isolates resistant to carbapenems were grown to the mid-log phase (OD600: 0.4-0.5). Cell pellets were collected and total RNA was extracted using Triazole reagent (Sigma, USA). RNA was purified and cDNA was synthesized using Quanti Tect Rev. Transcription Kit (Qiagen, Hilden, Germany) according to the manufacturer's instructions.
The expression level of both mexB and oprD was relatively normalized to the expression of rpoD (housekeeping gene). Overexpression of efflux pump MexAB-OprM was established when the expression of mexB was 2-3 folds its expression level in the wild isolate PAO1 [19][20]. The expression level of oprD was also calibrated relative to PAO1 [21].

Molecular characterization of P. aeruginosa
PFGE was performed as illustrated previously by Lopes et al., 2015 [22]. In brief, genomic DNA was digested with 50 U of SpeI (New England Biolabs, Massachusetts, USA). DNA fragments were separated by PFGE using a CHEF-DR III system. The running conditions include initial switch time of 2.2 seconds, running time for 22 hours, gradient of 6 volt/cm, angle of 120, temperature of 14 ºC and a final switch time of 54.2 seconds. Gels were stained with 1 µg/mL of ethidium bromide (Sigma Life Science) for 30 minutes and gel images were captured using UV gel image acquisition camera, Gel Doc XR (Bio-Rad Laboratories, Carlsbad, USA). PFGE pattern comparison is based on published procedures [23]. DNA relatedness was detected based on DICE coefficient with 4% optimization and 2% tolerance using BioNumerics Software (Applied Maths, Sint-Martens-Latem, Belgium).

Statistical analysis
The mean and standard deviation of the mean (SD; three replicates) were calculated using Microsoft Excel. All results were presented as mean ± SD. Statistical tests were applied as described under each experiment.

Results
Antimicrobial susceptibility test Ninety P. aeruginosa isolates were purified and identified biochemically. The isolates were obtained from different clinical sources including urine, pus swabs and sputum. P. aeruginosa isolates were collected from University Hospital (26 isolates), Chest Hospital (13 isolates), Urology and Nephrology Center (24 isolates), Burns and Cosmetics Center (11 isolates), and International Hospital (16 isolates). The antimicrobial susceptibility testing revealed that sixteen isolates (17.7%) were resistant to imipenem. All imipenem resistant isolates were also meropenem resistant except isolate number 53 ( Table 2). Most imipenem resistant isolates were resistant to other antimicrobials including gentamicin (56.25%), levofloxacin (81.25%), ceftazidime (62.5%) and amikacin (62.5%). Eight isolates (50%) were defined to be multiple drug resistant ( Table 2). The MICs of imipenem ranged from 32 to 2048 µg per mL and the MICs values of meropenem resistant isolates ranged from 16 to 256 µg per mL (Table 3).

MHT
Detection of carbapenemases by Modified Hodge test revealed that twelve isolates were carbapenemase producers ( Table 3).

Effects of efflux pump inhibitors on carbapenem resistance
When efflux pump inhibitors CCCP (40 µg per mL) and PAbN (50 µg per mL) were added to imipenem, a significant reduction in MICs of imipenem was detected (Table 3). Treating the bacteria with efflux pump inhibitors alone without imipenem showed no effect on bacterial growth. In addition, the efflux inhibitory effect of PAbN was more significant compared to CCCP.

Expression of efflux pump and outer membrane protein
RT-PCR analysis of mexB revealed that seven carbapenem-resistant isolates exhibited high expression level of mexB ( Table 3). Loss of outer membrane OprD was exhibited by seven isolates including sputum isolates number 13, 28 and 29, urine isolates number 41, 53, and 56 and isolate number 89 purified from wound (Table 3).   Molecular characterization of P. aeruginosa PFGE of SpeI-digested genomic DNA from 16 P. aeruginosa isolates resulted in 11 distinct pulsed-field profiles (PFPs) comprising 11-15 restriction fragments with 70% similarity. The 11 PFPs were identified as A, B, C1, C2, D, E1, E2, F, G1, G2 and H with two P. aeruginosa isolates belonging to related clusters C1 and C2, five isolates to E1 and E2 and two isolates to G1 and G2 (Figure 1). Three P. aeruginosa isolates belonged to clusters B, D and H with one isolate in each cluster. Interestingly, we found that the isolates with multiple carbapenemases genes (blaNDM-1, blaVIM-1, blaVIM-2), correspond to the profiles E1 and E2 (Table  4).

Discussion
Antimicrobial resistance is a significant driver of high mortality and morbidity associated with bacterial infections all over the world. Carbapenems have been used as the last choice for treatment of many bacterial infections [24]. The current study report carbapenem resistance in 17% of the tested isolates. The highest number of carbapenem-resistant isolates has been identified among the isolates from pus swab (7/17) ( Table 2). Moreover, eight carbapenem resistant isolates (50%) are multi-drug resistant to three classes of antimicrobial agents (Table 2). Overall, high resistance to extended-spectrum β-lactams and carbapenems were reported in P. aeruginosa worldwide [4]. A rapid dissemination of carbapenemases enzymes has been reported in different countries, including India, Pakistan, Australia, Germany, Belgium, Canada, Kenya, Sultanate of Oman, and United States [25][26].
The MICs of imipenem in this study were very high, reaching 2048 µg per mL (Table 3). Therefore, these studied P. aeruginosa isolates could have other resistance mechanisms. Overexpression of efflux proteins and decreased drug permeability may also affect the activity of carbapenems on P. aeruginosa [1,5]. A significant decrease in MIC of imipenem in combination with 50 µg per mL PAbN or 40 µg per mL CCCP compared to that of imipenem alone was determined. These results point out the contribution of the active efflux pump to the emergence of carbapenem-resistant P. aeruginosa. Moreover, realtime PCR analysis confirmed overexpression of genes involved in active efflux machinery in seven isolates (43.7%) with 3-300 folds overexpression of mexB relative to PAO1. Upregulation of mexB was observed in P. aeruginosa isolates with elevated MICs of imipenem (128-2048 µg/mL) and meropenem (64-256 µg/mL). These findings indicate the role of active efflux pump in carbapenem resistance. In previous research, overexpression of MexAB-OprM has been identified as the primary mechanism of meropenem resistance [40].
Moreover, both meropenem and imipenem utilize the OprD pathway for entry into Pseudomonas cell [5]. Therefore, reduced OprD expression also contributes to reduced susceptibility to carbapenems in P. aeruginosa [41]. In the current study, seven isolates (43.7%) had low OprD expression, five of them were highly resistant to imipenem (MIC 128-2048 µg/mL) and two were resistant to meropenem (MIC 64-256 µg/mL). In the study of Yi et al., 2006 [40], imipenem resistance is mainly mediated by OprD deficiency. In agreement with that, Kim and coauthors have recently reported that OprD inactivation is the most common mechanism of carbapenem resistance in P. aeruginosa isolated from Korean patients [42].
In this study, combined mechanisms of carbapenem resistance were recorded in most P. aeruginosa isolates. It was found that 56% of the resistant isolates exhibited two mechanisms of resistance to carbapenem and 25% of the resistant isolates possessed three mechanisms. Recently, the same observation of several resistance mechanisms to carbapenem was reported in P. aeruginosa isolates from Taiwan [43] and Spain [44].
Another significant observation in our study is the close genetic relation of isolates from different hospitals as indicated by PFGE (Figure 1). Isolates 85 and 89 belong to clone A and they were collected from Mansoura University Hospital and Burns and Cosmetics Center, respectively. Similarly, isolates 79 and 86 of clone F were isolated from Burns and Cosmetics and Mansoura International Hospital, respectively. Moreover, isolates 36 and 37 belong to clone E2 and they were isolated from University Hospital and Urology and Nephrology Center, respectively. These results suggest a possibility of microbial dissemination from one hospital to another rather than intrahospital transmission of these isolates.

Conclusion
In the present study, carbapenemases enzymes were identified in carbapenem resistant isolates with a high prevalence of blaVIM-1, blaVIM-2 and blaNDM-1. Most isolates harbored two or three carbapenemase genes. Induction of the efflux machinery and reduced expression of OprD are also involved in carbapenem resistance of P. aeruginosa. Moreover, the emergence of similar clones of carbapenem-resistant P. aeruginosa suggests microbial dissemination of the isolates with multiple carabapenemases. Hence, active surveillance of carbapenemase enzymes in addition to other carbapenem resistance mechanisms and enforcement of appropriate infection control precautions are obligatory to inhibit the dissemination of carbapenem-resistant P. aeruginosa infection.